Submit Manuscript

Journal of Applied Sciences

Year: 2007 | Volume: 7 | Issue: 23 | Page No.: 3750-3755
Facebook Twitter Digg Reddit Linkedin StumbleUpon E-mail

Article Trend

Total views 294

Search

Authors

G.R. Jahanshahloo

M. Afzalinejad

Keywords

Research Article

Quadratic Frontier for the Production Possibility Set in Data Envelopment Analysis

G.R. Jahanshahloo and M. Afzalinejad
Data Envelopment Analysis (DEA) applies simple-structured linear programs to evaluate the efficiency of similar Decision Making Units (DMUs). For large-scale problems with a huge number of DMUs, however, it is a computationally expensive method. Since in large-scale problems the piece-wise linear frontier of the Production Possibility Set (PPS) tends to be a smooth nonlinear surface, it is reasonable to consider such a nonlinear frontier in efficiency evaluations. In this study, a parametric DEA method is proposed to estimate a quadratic frontier for the PPS in constant returns-to-scale environment. Quadratic frontiers better fit the reality of large-scale problems. In addition, parametric approaches reduce the computations to a great extent. With a numerical example, the proposed approach is compared with the CCR model, which shows that the results of the proposed approach are good approximations, even for small-size problems.
PDF Fulltext XML References Citation

Leave a Comment

Your email address will not be published. Required fields are marked *